|Publication number||US7120213 B2|
|Application number||US 10/037,179|
|Publication date||Oct 10, 2006|
|Filing date||Oct 23, 2001|
|Priority date||Oct 27, 2000|
|Also published as||EP1202486A2, EP1202486A3, US7450672, US20020115463, US20060029169|
|Publication number||037179, 10037179, US 7120213 B2, US 7120213B2, US-B2-7120213, US7120213 B2, US7120213B2|
|Inventors||Alan Gatherer, Tarik Muharemovic, Everest W. Huang, Srinath Hosur|
|Original Assignee||Texas Instruments Incorporated|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (11), Non-Patent Citations (8), Referenced by (4), Classifications (24), Legal Events (3)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application claims the priority under 35 USC 119(e)(1) of copending U.S. provisional application number 60/244,043 filed on Oct. 27, 2000. This application contains subject matter related to subject matter disclosed in U.S. application Ser. No. 09/925,077 filed on Aug. 8, 2001, which is incorporated herein by reference.
The invention relates generally to wireless communications and, more particularly, to wireless communications that utilize turbo coding and transmit diversity.
Each of the documents listed below is referred to herein by the corresponding number enclosed in square brackets to the left of the document. Each of these documents is also incorporated herein by reference.
Coding and interleaving techniques are often used in wireless communication systems to improve the communication performance.
It is desirable in view of the foregoing to provide for improved performance in wireless communication systems that utilize turbo coding and transmit diversity.
According to the invention, coded bits and an interleaved version of the coded bits are separately modulated and transmitted. On the receiver side, a priori output probabilities produced by the probability generator are combined and then input to a SISO decoder, and combined a posteriori output probabilities produced by the SISO decoder are split and then fed back to the probability generator. This advantageously permits the probability generator to produce an improved estimate of the received symbols.
Referring again to
Z t =αX t +βY t +n t, (1)
where nt represents noise in the wireless communication channels. At 13, the probability generator 15 produces, for all possible values CX that Xt can assume at time t, the following probability
P(X t =C X |Z t =C Z) (2)
Expression (2) above represents the probability that Xt=CX given that the received symbol or bit value Zt=CZ. At 14, the probability generator 15 produces similar probabilities for all possible values CY of Yt, namely
P(Y t =C Y |Z t =C Z) (3)
Taking the probability defined in Expression (2) above as an example, and applying Bayes' Rule, Expression (2) can be written as follows
P(Z t =C Z |X t =C X)P(X t =C X) /P(Z t =C Z) (4)
In practice, for an iterative loop, the probability given by Expression (2) is generated under the assumption that nothing is known in advance about the statistics of Xt. This is called the extrinsic probability and ensures that only “new” information is used to generate data that will be fed back. Therefore, P(Xt=CX) can be eliminated from Expression (4). The denominator of Expression (4) can also be eliminated because it merely represents the probability that Zt=CZ at time t, which is merely a constant value that operates only as a scaling factor. Thus, eliminating the aforementioned extrinsic factor and the aforementioned scaling factor from Expression (4) leaves
P(Z t =C Z |X t=CX) (5)
Using known probability theory, Expression (5) can be rewritten as follows
Referring again to Equation (1) above, the leftmost probability of Expression (6) can be rewritten as follows
P(n t =C Z −αC X −βC Y) (7)
Substituting Expression (7) into Expression (6) gives
Thus, Expression (2) above can be rewritten as Expression (8) above.
The noise nt in Expression 8 can be modeled as a Gaussian random variable, and the fading parameters α and β can be readily estimated. Thus, given that the received symbol Zt=CZ is known, values of the leftmost probability in Expression 8 can be easily calculated for all possible values of CX and CY. The values of the rightmost probability of Expression (8) are provided according to the invention as the a posteriori output probabilities from a SISO decoder, as described in more detail below.
Using reasoning analogous to that given above for rewriting Expression (2) as Expression (8), Expression (3) above can be rewritten as follows
As mentioned above with respect to Expression (8), the leftmost probability of Expression (9) can be easily calculated for a known value of CZ and all possible values of CX and CY. Also analogous to the discussion of Expression (8) above, the values of the rightmost probability of Expression (9) are provided according to the invention as a posteriori output probabilities of a SISO decoder.
Referring now to
The outputs 45 and 46 provide a priori output probabilities to the SISO decoders 35 and 36. The decoder 35 operates with respect to Xt and the decoder 36 operates with respect to Yt. The SISO decoders 35 and 36 use their respective a priori output probabilities to produce respective a posteriori input probabilities. The a posteriori input probabilities produced by SISO decoder 35 are interleaved at 38 (corresponding to the interleaver in the transmitter 11) and the results are provided as a priori input probabilities to the SISO decoder 36. Similarly, the a posteriori input probabilities produced by the SISO decoder 36 are de-interleaved at 37 (again corresponding to the interleaver of the transmitter 11) and the results are provided as a priori input probabilities to the SISO decoder 35. The a posteriori input probabilities produced by the SISO decoder 35 are also provided to a decision maker which can use conventional techniques to decide the input symbol (as seen by the corresponding coder 16) based on the a posteriori input probabilities.
The output probabilities provided to (a priori) and produced by (a posteriori) the SISO decoder 35 represent respective probabilities that the symbol Xt as output from the convolutional coder 16 has respective ones of a plurality of possible values. Similarly, the input probabilities provided to (a priori) and produced by (a posteriori) SISO decoder 35 represent respective probabilities that the symbol that was input to the convolutional coder 16 to produce Xt has respective ones of a plurality of possible values. The SISO decoder 36 functions analogously with respect to the symbol Yt and the convolutional coder 17. Each SISO decoder uses the a priori probabilities (input and output) provided thereto together with knowledge of the trellis used by the corresponding convolutional coder to produce corresponding a posteriori probabilities (output and input). In some embodiments, each coder 16 and 17 uses the same trellis.
The combiner 43 receives at 47 the a posteriori output probabilities produced by SISO decoder 35, and the combiner 44 receives at 48 the a posteriori output probabilities produced by SISO decoder 36. The values received at 47 represent the values of the rightmost probability in Expression (9) and the values received at 48 represent the values of the rightmost probability in Expression (8). The combiner 43 operates to combine the values that it receives at 49 and 47 in the manner shown in Expression (8), namely multiplying the values together and summing the resulting products over all possible values of CY. Similarly, the combiner 44 combines the values that it receives at 49 and 48 as shown by Expression (9) above, namely multiplying the values together and summing the resulting products over all possible values of CX. The combiner 43 produces at 46 the values of the probability shown in Expression (3), and the combiner 44 produces at 45 the values of the probability shown in Expression (2).
It should be clear that the probability generator 34 can easily account for the scaling factor described above with respect to Expression (4) by suitably normalizing the probability values that it generates, although such normalizing is not explicitly shown in the drawings.
At 52, SISO 35 uses the initial a priori output probabilities to produce a posteriori input probabilities. At 53, interleaving is applied to the a posteriori input probabilities from SISO 35. At 54, SISO 36 uses the initial (for the first iteration) a priori output probabilities and the interleaved a posteriori input probabilities of SISO 35 to produce a posteriori input and output probabilities. At 55, de-interleaving is applied to the a posteriori input probabilities from SISO 36. At 56, the a posteriori output probabilities from SISO 36 are used to produce a priori output probabilities for SISO 35. At 57, the SISO 35 uses its a priori output probabilities and the de-interleaved a posteriori input probabilities of SISO 36 to produce a posteriori input and output probabilities. At 58, the a posteriori output probabilities from SISO 35 are used to produced a priori output probabilities for SISO 36. The operations at 53–58 are then repeated for any desired number of iterations.
The combiner 112 is operable for combining the probability values at 46 with the probability values output by the de-interleaver 110. In some exemplary embodiments, the combiner is simply a multiplier which multiplies the input probability values by one another. The combiner 112 thus outputs combined a priori output probability values which represent combinations of the a priori output probability values input to the combiner 112. The combined a priori output probability values at 114 are provided to a SISO decoder 111. The SISO decoder 111 uses the combined a priori output probability values 114 to produce combined a posteriori input and output probabilities. The combined a posteriori input probabilities are provided to a decision maker which decides the symbol values, and the combined a posteriori output probabilities are provided at 115 to a splitter 113.
The splitter 113 is operable for splitting each of the combined a posteriori output probability values at 115 into its constituent probability values. The splitter output values 116, corresponding to probability values 46, are provided to input 48 of the probability generator 34, and the splitter output values 117, corresponding to the probability values at 45, are applied to an interleaver 92 (same as in
The decision maker can also utilize, for example, a splitter such as shown at 113 to split each of the combined a posteriori input probabilities into its constituent probability values. These constituent probability values can then be used in conventional fashion to make the symbol decisions.
It will be apparent to workers in the art that any wireless communication system that utilizes a space-time turbo code, or any kind of turbo code, can benefit from the present invention. Advantageously, the added complexity of the a posteriori output probability feedback loops is relatively small compared to the complexity of a SISO block. It will also be apparent to workers in the art that the embodiments of
Although exemplary embodiments of the invention are described above in detail, this does not limit the scope of the invention, which can be practiced in a variety of embodiments.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5883907||May 1, 1997||Mar 16, 1999||Motorola, Inc.||Asymmetrical digital subscriber line (ADSL) block encoder circuit and method of operation|
|US6307901 *||Apr 24, 2000||Oct 23, 2001||Motorola, Inc.||Turbo decoder with decision feedback equalization|
|US6671338 *||Nov 10, 1999||Dec 30, 2003||Hughes Electronics Corporation||Combined interference cancellation with FEC decoding for high spectral efficiency satellite communications|
|US6757701 *||Aug 3, 2001||Jun 29, 2004||Motorola, Inc.||Apparatus and method for implementing a linearly approximated log map algorithm|
|US6763494 *||Dec 19, 2002||Jul 13, 2004||Comtech Telecommunications Corp.||Turbo product code decoder|
|US6789218 *||Oct 30, 2002||Sep 7, 2004||Icoding Technology, Inc.||High spread highly randomized generatable interleavers|
|US6891897 *||Oct 27, 2000||May 10, 2005||Nortel Networks Limited||Space-time coding and channel estimation scheme, arrangement and method|
|US20020041646 *||Aug 8, 2001||Apr 11, 2002||Huang Everest W.||Using SISO decoder feedback to produce symbol probabilities for use in wireless communications that utilize turbo coding and transmit diversity|
|US20020110206 *||Dec 13, 2001||Aug 15, 2002||Neal Becker||Combined interference cancellation with FEC decoding for high spectral efficiency satellite communications|
|EP0982870A1||Aug 21, 1998||Mar 1, 2000||Lucent Technologies Inc.||Multicode CDMA system using iterative decoding|
|WO1998043373A1||Mar 26, 1998||Oct 1, 1998||Amir Bigloo||Systems and methods of channel coding and inverse-multiplexing for multi-carrier cdma systems|
|1||"Combined Multilevel Turbo-Code With MR-Modulation", L. Papke, et al., Communications-Gateway to Globalization. Proceedings of the Conference on Communications. Seattle, Jun. 18-22, 1995, Proceedings of the Conference on Communications (ICC), New York, IEEE, US, vol. 2, Jun. 18, 1995, 668-672, XP000533098, ISBN:9-7803-2487-0.|
|2||"Improved Space-Time Codes Using Serial Concatenation", Xiaotong Lin, et al., IEEE Communications Leters, IEEE Service Center, Piscataway, NJ, US, vol. 4, No. 7, Jul. 20000, pp. 221-223, XP000958655, ISSN: 1089-7798.|
|3||A.R. Hammons and H.E. Gamal, "On the Theory of Space-Time Codes for PSK Modulation", In IEEE Transactions on Information Theory, vol. 2, No. 2, pp. 524-542, Mar. 2000.|
|4||V. Tarokh, N. Seshadri, and A.R. Calderbank, "Space-Time Codes for High Data Rate Wireless Communication: Performance Criterion and Code Construction", In IEEE Transactions on Information Theory, vol. 44, No. 2, pp. 744-765, Mar. 1998.|
|5||X. Li and J.A. Ritcey, "Bit Interleaved Coded Modulation With Iterative Decoding", in IEEE ICC, vol. 2, pp. 858-863, Jun. 1999.|
|6||X. Li and J.A. Ritcey, "Bit-interleaved Coded Modulation With Iterative Decoding Using Soft Feedback", Electronic Letters, vol. 34, pp. 942-943, Apr. 3, 1998.|
|7||X. Li and J.A. Ritcey, "Bit-Interleaved Coded Modulation with Iterative Decoding", IEEE Communication Letters, vol. 1, pp. 169-171, Nov. 1997.|
|8||Y. Liu, M.P. Fitz, and O.Y. Takeshita, "Qpsk Space-Time Turbo Codes", In IEEE ICC, Jun. 2000.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7664186 *||May 11, 2004||Feb 16, 2010||Infineon Technologies Ag||Channel decoding for multicarrier signal transmission by means of DC-offset and carrier-frequency offset-dependent weighting of reliability information|
|US7680219 *||May 24, 2005||Mar 16, 2010||Trident Microsystems (Far East) Ltd.||Apparatus and method for decoding a bit sequence from QPSK or QAM symbols|
|US20040228418 *||May 11, 2004||Nov 18, 2004||Lars Brotje||Channel decoding for multicarrier signal transmission by means of DC-offset and carrier-frequency offset-dependent weighting of reliability information|
|US20050259765 *||May 24, 2005||Nov 24, 2005||Stefan Keller||Apparatus and method for decoding a bit sequence from QPSK or QAM symbols|
|U.S. Classification||375/347, 455/69, 375/341, 714/701, 375/346, 375/265, 708/277, 455/522|
|International Classification||H03M13/27, H04L1/02, H03M13/45, H04B7/10, H04L1/06, H04L1/00|
|Cooperative Classification||H04L1/0071, H04L1/06, H04L1/005, H04L1/0066, H04L1/0618|
|European Classification||H04L1/06, H04L1/00B7V, H04L1/06T, H04L1/00B5E5, H04L1/00B7K3|
|Oct 23, 2001||AS||Assignment|
Owner name: TEXAS INSTRUMENTS INCORPORATED, TEXAS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:GATHERER, ALAN;MUHAREMOVIC, TARIK;HUANG, EVEREST W.;AND OTHERS;REEL/FRAME:012455/0506;SIGNING DATES FROM 20011016 TO 20011022
|Mar 23, 2010||FPAY||Fee payment|
Year of fee payment: 4
|Mar 26, 2014||FPAY||Fee payment|
Year of fee payment: 8